1. This site uses cookies. By continuing to use this site, you are agreeing to our use of cookies. Learn More.

Genetic Algorithms

Discussion in 'Scripting & Programming' started by Fergal1982, Sep 17, 2007.

  1. Fergal1982

    Fergal1982 Petabyte Poster

    I picked up a book on Genetic Algorithms (GAs) at the weekend and, whilst I've only read part of the first chapter (and have been confused by a decent bit of it), I'm finding it quite interesting so far.

    For those of you who arent aware (most I think), I actually went to Uni to study Genetics, but dropped (:oops: ) out before completing my degree. So this looked like an interesting opportunity to marry what I enjoyed about Genetics, with programming.

    I've found it interesting to the extent that I found I couldn't sleep last night and, for the couple of hours it took to get to sleep, my mind was thinking about GAs over and over. I found myself stuck on the first example given in the book (Basically, trying to get a 6-character string equal to '111111'). I was even going over in my head how I could incorporate another type of mutation into the mix (insertion - where 'DNA' is inserted randomly into a portion of the existing Genetic data. I actually got to the extent where I thought up a system of determining the fitness of each candidate string in a generation given that the length of the string would vary (Number of positions matching the target string, divided by the length of the string or the length of the target string, whichever is greater).

    Its a simple example of GAs, but one that I found I could (at least a little), get my head around. Not, however, something you want running through your head at two in the morning when its not helping you sleep (grrr).

    Has anyone else covered the subject of GAs at all? What did you think of the subject? And can you recommend any good resources?
    Certifications: ITIL Foundation; MCTS: Visual Studio Team Foundation Server 2010, Administration
    WIP: None at present
  2. dmarsh

    dmarsh Terabyte Poster

    I've come across various stuff in various jobs, The Kalman Filters, Neural Networks, Scorecards, Rules Engines, Control Algorithms, read basic popular science books on GA, Artificial Life never messed with them though.

    Aritificial Life certainly looks more promising to me than the traditional AI camp.
  3. Mathematix

    Mathematix Megabyte Poster

    I specialised in AI for my final year at University and studied Neural Nets and Genetic Algorithms in some depth, even though I have forgotten a fair old bit now.

    Your training in genetics will go a long way in studying GAs. The best way to see it is that you are manipulating 1s and 0s rather than ACGT when studying the evolutionary algorithms. Not that this necessariliy makes the subject any easier, since you are borrowing techniques from nature to improve 'fitness' for decision making and other problems involving chance.

    We studied from lectures and books rather than the web, so cannot point you to any web resources. Genetic Algorithms in Search, Optimization and Machine Learning was our main course material and is an excellent book. I would also suggest that you brush up on Fuzzy Logic/Systems as a foundation for any further study. :)
    Certifications: BSc(Hons) Comp Sci, BCS Award of Merit
    WIP: Not doing certs. Computer geek.

Share This Page